Autonomní kontrola stavu přejezdových závor pomocí počítačového vidění

Abstract

This diploma thesis deals with the problem of autonomous control of level crossing barrier using computer vision. The introduction explains the motivation to create this work and it presents the problem to be dealt with. The theoretical part describes the methods used in the design of the detector. Next section summarizes what and why the field of object recognition in images is used for and examples are given of applications in this field. Two approaches have been designed and implemented to detect the level crossing barriers. The main part of the thesis describes their principle and development. Measuring the success of a created detectors is done in chapter of experiments. The conclusion summarizes the experience gained. Created detectors are compared in this chapter too.